As a channel-prosthesis, the mirror can be a source of perceptive de-ception, just as any other prosthesis. I enter a room and seem to see a man coming toward me, and then I realize it is my image reflected in a mirror. This image ‘standing for something else’, albeit temporarily, might induce us to perceive the shadow of a semiotic phenomenon. But it is a perceptive deception, as I can well have without mirrors, as when I take dross for gold or see things that are not there. Similarly, decep-tions can be created by presenting things which are not mirrors as being mirrors. In a film by the Marx Brothers there is a scene where Groucho is looking at himself in a mirror, but the mirror is not a mirror; it is an empty frame behind which Harpo is awkwardly (and with funny effects) trying to imitate Groucho’s gestures. This phenomenon of lying about mirrors of course has nothing to do with mirror images. No doubt the deceiver’s performance has something to do with fiction, with signification, with lying through signs, but all this does not concern the nature of the mirror image. This will come up again (7.12) when we deal with a semiotics of the mise-en-scene, which might apply to the use of mirrors as channels.
7.7. Absolute icons
We have said that a catoptric prosthesis extends an organ range of action and supplies the organ with the same stimuli it would receive if it could act right where the prosthesis extends its range. In this sense, the mirror provides me with an absolute double of the stimulating field. We might quite naively say that the mirror provides me with an icon of the object —if we define the icon as an image having (if the properties of its denotatum. But catoptric experience tells me that (if any sign called icon and endowed with these properties exists) a catoptric absolute icon is not an icon but a double (see Eco 1975, 3.4.7). At the macroscopic level of my perceptive experience and to the practical purpose it must serve, the sheet of paper I am writing on is the double of the sheet I have just filled.
But this is not a good reason to consider the former a sign of the latter. You may argue that the mirror image is not to its object as the former sheet is to the latter. But you should not forget that the mirror image is not a double of its object but, rather, a double of the stimulat-ing field one could have access to if one looked at the object instead of looking at its mirror image. The fact that the mirror image is a most peculiar case of double and has the traits of a unique case explains why mirrors inspired so much literature; this virtual duplication of stimuli (which sometimes works as if there were a duplication of both my body as an object and my body as a subject, splitting and facing itself), this theft of an image, this unceasing temptation to believe I am someone else, makes man’s experience with mirrors an absolutely unique one, on the threshold between perception and signification.
And it is precisely from this experience of absolute iconism that the dream of a sign having the same characteristics arises. This is why men draw (and produce the signs which are precisely defined as iconic): they draw to achieve with-out mirrors what mirrors allow them to achieve. But the most realistic drawing does not show all the characteristics of absolute duplication as a mirror does (besides having a different relation of dependence to its object).
Man’s experience with mirrors may then explain the emergence of a notion like the (semiotic) one of iconism but is not explained by it.
However, the mirror as a threshold phenomenon may lend itself to a number of operations making it even more ‘threshold’. I can, in fact, reduce the absolute iconism of mirror images, and smoked mirrors are an excellent example of this technique. The mirror almost becomes a re-ducing prosthesis.
Let us imagine a mirror made of horizontal strips of reflecting material with thin opaque strips in between. The virtual image I see is obviously incomplete. At the level of perceptive reconstruction, the result may nevertheless be excellent, with varying degrees of efficiency depending on the thickness of opaque strips. If we imagine opaque strips of a rea-sonable thickness, even though the reflected image is not mine (be-cause, of course, I know a lot about my image, and the reconstruction of the perceptum may in this case be affected by previous information), I can satisfactorily perceive the reflected object.
This, of course, does not ex-clude that some elements of interpretation (although very slight ones) do come into play. Such interpretation, however, also affects the perception of objects in the world around us. Darkness, the presence of opaque obstacles, fog are all ‘noises’ in the channel diminishing sensory data and requiring interpretative efforts in order to achieve the (often conjectural) formation of a perceptum. If these conjectural and interpretive efforts are to be taken as semiosic, then semiosis creeps into any aspect of our relation to the surrounding world. But, even if we take this for granted, we should not conclude that any aurorally semiosic process is productive and interpretive of signs. If also mirrors impose semiosic processes, one thing remains to be defined, that is, in which sense these processes do not lead to the production, interpretation, and use of signs.
7.8. Mirrors as rigid designators
The mirror has a peculiar characteristic. As long as I look at it, it gives me back my facial features, but if I mailed a mirror which I have long looked at to my beloved, so that she may remember my looks, she could not see me (and would instead see herself)·
The self-evident datum I have just highlighted deserves some thought. If we compared mirror images to words, they would be like personal pronouns: like the pronoun /I/, meaning «Umberto Eco» if I pronounce it, and someone else if someone else does so. I may, how-ever, happen to find a message in a bottle reading «I was shipwrecked in the Juan Fernandez islands»; it would be clear to me that someone (someone who is not myself) was shipwrecked. But, if I find a mirror in a bottle, after taking it out with considerable effort, I would always see myself in it, whoever may have sent it as a message. If the mirror ‘names’ (and this is clearly a metaphor), it only names a concrete object, it names one at a time, and it always names only the object standing in front of it. In other words, whatever a mirror image may be, it is deter-mined in its origins and in its physical existence by an object we shall call the image referent.
In an extreme attempt to find one more relation between mirror im-ages and words, we might compare mirror images to proper names. If in a crowded station I would shout John! I am likely to see a great many people turning around —which allowed many to say that proper names have a direct relation to their bearers. Yet, if someone looking out of the window would say Look, there comes John!, inside the room and without knowing John, I would know that the other saw (or says he saw) a male human being (provided he is making an appropriate use of language). If this is so, then even proper names do not refer directly to an object whose presence determines the proper name utterance. Not only could my comrade lie, and mention John when John is not there, but the lin-guistic expression John first and foremost refers me to a general content. So much so that, if someone would eventually decide to christen his newborn daughter John, I would tell him that he is making an inapprop-riate use of current onomastics, since John usually names males.
Therefore, there is a difference between a mirror image and a proper name, in that a mirror image is an absolute proper name as it is an absolute icon. In other words, the semiotic dream of proper names being im-mediately linked to their referent (just like the semiotic dream of an image having all the properties of the object they refer to) arises from a sort of catoptric nostalgia. There is actually a theory of proper names as rigid designators (Kripke 1972) by which proper names could not be mediated by definite descriptions (like John is the fellow who . . . ) but could undergo counterf actual exercises (like Would John still be John, were he not the fellow who . . .?). An unbroken chain of designations, called a ‘causal’ chain, links them to an original object they were allocated to by a sort of initial ‘baptism’.
Now, it is mirrors which allow us to imagine this kind of situation. Let
